# Fennie - AI-Powered Academic Tutoring Platform

## Project Overview
Fennie is a closed-source SaaS application that provides AI-powered academic tutoring while maintaining strict academic integrity standards. The platform guides students through learning processes rather than providing direct answers, ensuring genuine understanding and ethical academic practices.

## Core Philosophy
- **Academic Integrity First**: Never provides direct answers to homework or assignments
- **Guided Learning**: Uses step-by-step problem-solving approaches
- **Ethical AI**: Cannot be circumvented through clever prompting
- **Student-Centered**: Focuses on developing genuine understanding

## Technology Stack

### Frontend
- **Next.js 15** with App Router for routing and SSR
- **React 19** for UI components
- **Tailwind CSS 4** for styling with dark/light theme support
- **ShadCN/UI** for pre-built accessible components
- **Zustand** for client-side state management
- **TypeScript** for type safety

### Backend
- **Next.js API Routes** for serverless backend functions
- **Prisma ORM** for type-safe database operations
- **Supabase Auth** for secure authentication (OAuth + email/password)
- **OpenAI SDK** for AI integration (GPT-4o-mini for tutoring)
- **MCP (Model Context Protocol)** server for intelligent note management

### Database
- **PostgreSQL** with pgvector extension for vector embeddings
- **Supabase** for managed database hosting
- **Prisma** for schema management and migrations

### AI & ML
- **OpenAI GPT-4o-mini** for chat completions and tutoring
- **text-embedding-3-small** for semantic search
- **Vector embeddings** for context-aware responses
- **User memory system** with confidence tracking

### Infrastructure
- **Vercel** for deployment and hosting
- **Google OAuth** for authentication
- **KaTeX** for mathematical equation rendering
- **Mermaid** for interactive diagram generation

## Key Features

### Authentication & User Management
- OAuth authentication (Google)
- Email/password authentication with Supabase
- Password reset functionality
- Secure session management with JWT

### AI Chat Interface
- Real-time chat with AI tutor
- Context-aware responses using user notes
- Chat history and persistence
- Multiple conversation management
- Academic integrity enforcement
- Mathematical equation support (KaTeX)
- Code syntax highlighting
- Mermaid diagram rendering

### Knowledge Management
- **Notes**: Create, edit, and organize study notes with markdown support
- **Documents**: Upload and process PDF/DOCX files for AI context
- **User Memory**: AI tracks student knowledge and learning progress
- **Vector Search**: Semantic search across notes and documents
- **Tags**: Organize content with tagging system

### Study Tools
- **Flashcards**: Generate and study flashcard sets
- **Quizzes**: Create interactive quizzes from content
- **Study Guides**: AI-generated study materials
- **Diagnostic Questioning**: Assess baseline understanding

### Collaboration & Sharing
- Public sharing of notes and chats via secure tokens
- Import/export functionality for collaboration
- Blog system for educational content

### SEO & Performance
- Server-side rendering (SSR) and static generation (SSG)
- Comprehensive meta tags and structured data
- XML sitemap generation
- Core Web Vitals optimization
- OpenGraph image generation

## Project Structure

```
aintegrity/
├── src/
│   ├── app/                    # Next.js App Router pages
│   │   ├── (legal)/           # Legal pages (privacy, terms)
│   │   ├── api/               # API routes
│   │   │   ├── chat/          # Chat endpoints
│   │   │   ├── notes/         # Notes CRUD
│   │   │   ├── documents/     # Document processing
│   │   │   ├── flashcards/    # Flashcard management
│   │   │   ├── quizzes/       # Quiz functionality
│   │   │   └── user-memory/   # User knowledge tracking
│   │   ├── auth/              # Authentication pages
│   │   ├── blog/              # Blog system
│   │   ├── dashboard/         # Main application interface
│   │   ├── shared/            # Public sharing pages
│   │   └── lib/               # Utility libraries
│   ├── components/            # React components
│   │   ├── ui/                # ShadCN/UI components
│   │   ├── chat/              # Chat interface components
│   │   ├── notes/             # Note management components
│   │   └── shared/            # Shared components
│   ├── contexts/              # React contexts
│   ├── hooks/                 # Custom React hooks
│   └── types/                 # TypeScript type definitions
├── prisma/                    # Database schema and migrations
├── docs/                      # Comprehensive documentation
├── public/                    # Static assets
└── scripts/                   # Build and utility scripts
```

## Database Schema

### Core Models
- **User**: User accounts with Supabase integration
- **Note**: User knowledge base with vector embeddings
- **Chat/Message**: Conversation history
- **Document**: Uploaded files with extracted text
- **Quiz/Flashcard**: Study tools
- **UserMemory**: AI knowledge tracking system

### Key Relationships
- User → Notes (one-to-many)
- User → Chats (one-to-many)
- Chat → Messages (one-to-many)
- User → Documents (one-to-many)

## API Design

### Authentication
- `GET/POST /auth/*` - Supabase Auth handlers
- OAuth and email/password flows

### Core Features
- `POST /api/chat` - AI tutoring conversations
- `GET/POST/PUT/DELETE /api/notes` - Note management
- `POST /api/documents` - Document upload and processing
- `GET /api/search` - Semantic search across content
- `GET/POST /api/user-memory` - Knowledge tracking

## Development Guidelines

### Code Style
- Use outline variant for all buttons
- Implement proper dark/light theme support
- Follow academic integrity principles in AI responses
- Use TypeScript for type safety
- Implement proper error handling and loading states

### Security
- Row Level Security (RLS) policies in Supabase
- JWT token validation for API routes
- Secure file upload handling
- Input sanitization and validation

### Performance
- Server-side rendering for SEO
- Vector embeddings for fast semantic search
- Optimized database queries with Prisma
- Image optimization and lazy loading

## Environment Variables
- `NEXT_PUBLIC_SUPABASE_URL` - Supabase project URL
- `NEXT_PUBLIC_SUPABASE_ANON_KEY` - Supabase anonymous key
- `SUPABASE_SERVICE_ROLE_KEY` - Supabase service role key
- `OPENAI_API_KEY` - OpenAI API key
- `DATABASE_URL` - PostgreSQL connection string

## Deployment
- **Platform**: Vercel with serverless functions
- **Database**: Supabase managed PostgreSQL
- **Build Process**: Includes Prisma client generation
- **Environment**: Production optimized with analytics

## Brand Guidelines
- **Primary Color**: #1f1a14 (dark brown)
- **Theme**: Dark/light mode support
- **Typography**: Clean, academic-focused design
- **Social**: @FennieAi (Twitter), @fennie.ai (Instagram)

This codebase represents a production-ready SaaS application focused on ethical AI tutoring with comprehensive features for academic support while maintaining the highest standards of academic integrity.